Olfactory electroencephalogram signal recognition system and method

Abstract

The invention discloses an olfactory electroencephalogram signal recognition system and method. The system comprises an olfactory stimulation generation module, an electroencephalogram signal acquisition module, a near-infrared signal acquisition module, a synchronous control module and a data analysis processing module, the smell stimulation generation module is used for generating smell under the control of the synchronous control module in response to an input instruction of a user; an electroencephalogram signal acquisition module is used for acquiring an electroencephalogram signal of a user; acquiring a near-infrared signal of a user by using a near-infrared signal acquisition module; integrating the electroencephalogram signal and the near-infrared signal to obtain an olfactory electroencephalogram signal data set; and processing the olfactory electroencephalogram signal data set by using a data analysis processing module to obtain an olfactory electroencephalogram signal recognition result. The electroencephalogram signal and the near-infrared signal can be collected at the same time, fusion of multi-mode information is achieved, and the perception and cognition process of the brain to olfactory stimulation can be known more comprehensively.

Description

Technical Field

[0001] The present invention relates to the technical field of neuroscience and biomedical engineering, and in particular to an olfactory electroencephalogram signal recognition system and method. Background Art

[0002] Electroencephalogram (EEG) and near-infrared spectroscopy (NIRS) are important means of studying brain function and cognitive processes. EEG can reflect the electrical activity of brain neurons and has high temporal resolution, but relatively low spatial resolution. Near-infrared spectroscopy can measure the concentration changes of oxygenated hemoglobin and deoxygenated hemoglobin in brain tissue, thereby reflecting the hemodynamic changes of the brain, and has good spatial resolution.

[0003] In olfactory research, understanding the brain's neural response to odor stimulation is crucial to revealing the mechanisms of olfactory perception and cognition. However, existing olfactory EEG event-related potential acquisition devices and methods often have the following shortcomings:

[0004] 1. Collecting EEG signals or near-infrared signals alone cannot obtain information from both modalities at the same time, making it difficult to comprehensively and deeply study the brain's response to olfactory stimulation.

[0005] 2. The synchronization performance of the acquisition device is poor, resulting in inaccurate temporal matching between EEG signals and near-infrared signals, affecting the reliability of data analysis and results.

[0006] 3. The device is not comfortable and portable, which may cause discomfort or non-cooperation to the subjects, thus affecting the quality of the experimental data.

[0007] What is the "gold standard" for olfactory assessment, the predictive indicators for recovery of olfactory function, and the changes in the olfactory nerve and olfactory cortex are all unclear issues that need to be answered urgently. Summary of the invention

[0008] The technical problem to be solved by the present invention is to provide an olfactory EEG signal recognition system and method, which can simultaneously and accurately collect EEG signals and near-infrared signals, and provide more comprehensive and reliable data for olfactory neuroscience research.

[0009] In order to solve the above technical problems, the first aspect of the embodiment of the present invention discloses an olfactory electroencephalogram signal recognition system, the system comprising an olfactory stimulation generation module, an electroencephalogram signal acquisition module, a near-infrared signal acquisition module, a synchronization control module and a data analysis and processing module;

[0010] The olfactory stimulation generation module is data-connected to the synchronization control module and is used to generate odor;

[0011] The EEG signal acquisition module is data-connected with the synchronization control module and the data analysis and processing module, and is used to acquire the user's EEG signal and send the EEG signal to the data analysis and processing module;

[0012] The near infrared signal acquisition module is data-connected with the synchronization control module and the data analysis and processing module, and is used to acquire the user's near infrared signal and send the near infrared signal to the data analysis and processing module;

[0013] The synchronization control module is data-connected with the olfactory stimulation generation module, the EEG signal acquisition module and the near-infrared signal acquisition module for data synchronization;

[0014] The data analysis and processing module is data-connected to the EEG signal acquisition module and the near-infrared signal acquisition module, and is used to process the EEG signal and the near-infrared signal to obtain an olfactory EEG signal recognition result.

[0015] As an optional implementation, in the first aspect of the embodiment of the present invention, the olfactory stimulation generation module includes an odor generator and an odor delivery pipeline;

[0016] The odor generator is data-connected to the odor delivery pipeline and the synchronous control module, and is used to generate different odors under the control of the synchronous control module, and release the odors through the odor delivery pipeline;

[0017] The odor delivery pipeline is data-connected with the odor generator for outputting the odor.

[0018] As an optional implementation, in the first aspect of the embodiment of the present invention, the synchronization control module includes an input unit, an odor control unit and a synchronization control unit;

[0019] The input unit is data-connected to the odor control unit and is used to input user instructions;

[0020] The odor control unit is data-connected to the odor generator, and is used to control the odor generator to produce different odors according to the user instructions;

[0021] The synchronization control unit is data-connected with the odor generator, the EEG signal acquisition module and the near-infrared signal acquisition module for data synchronization.

[0022] A second aspect of an embodiment of the present invention discloses a method for identifying olfactory electroencephalogram signals, the method comprising:

[0023] S1, in response to a user's input instruction, under the control of a synchronous control module, using an olfactory stimulation generating module to generate an odor;

[0024] S2, collecting the user's EEG signal using an EEG signal collection module;

[0025] S3, using a near infrared signal acquisition module to collect a near infrared signal of the user;

[0026] S4, integrating the EEG signal and the near infrared signal to obtain an olfactory EEG signal data set;

[0027] S5, using a data analysis and processing module to process the olfactory EEG signal data set to obtain an olfactory EEG signal recognition result.

[0028] As an optional implementation, in the second aspect of the embodiment of the present invention, the use of the data analysis and processing module to process the olfactory EEG signal data set to obtain the olfactory EEG signal recognition result includes:

[0029] S51, extracting features from the EEG signal to obtain EEG signal feature information;

[0030] S52, extracting features from the near infrared signal to obtain feature information of the near infrared signal;

[0031] S53, performing feature fusion on the EEG signal feature information and the near infrared signal feature information to obtain fused feature information;

[0032] S54, using the fusion feature information, training the olfactory EEG signal recognition model to obtain an optimized olfactory EEG signal recognition model;

[0033] S55, using the optimized olfactory electroencephalogram signal recognition model, processing the electroencephalogram signal and the near-infrared signal to be processed to obtain an olfactory electroencephalogram signal recognition result.

[0034] As an optional implementation, in the second aspect of the embodiment of the present invention, the extracting features of the EEG signal to obtain EEG signal feature information includes:

[0035] S511, preprocessing the EEG signal to obtain a preprocessed EEG signal;

[0036] S512, performing first feature extraction on the preprocessed EEG signal to obtain first feature information;

[0037] S513, performing second feature extraction on the preprocessed EEG signal to obtain second feature information;

[0038] S514, extracting a third feature from the preprocessed EEG signal to obtain third feature information;

[0039] S515: Fusing the first feature information, the second feature information and the third feature information to obtain EEG signal feature information.

[0040] As an optional implementation manner, in the second aspect of the embodiment of the present invention, the calculation formula of the first feature information is:

[0041]

[0042] Among them, H1 is the first feature information, , the preprocessed EEG signal is X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

[0043] As an optional implementation manner, in the second aspect of the embodiment of the present invention, the calculation formula of the second characteristic information is:

[0044]

[0045] Among them, H2 is the second characteristic information, L is the preset length, j = 1, 2, ..., L, δ j is the singular value of matrix D, matrix The preprocessed EEG signal is X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

[0046] As an optional implementation manner, in the second aspect of the embodiment of the present invention, the calculation formula of the third characteristic information is:

[0047]

[0048] Among them, H3 is the third characteristic information, S i For x i The Fourier transform of i For S i The center frequency, X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

[0049] As an optional implementation manner, in the second aspect of the embodiment of the present invention, the feature extraction of the near infrared signal to obtain the near infrared signal feature information includes:

[0050] S521, using a time-frequency spectrum analysis model, performing time-frequency processing on the near-infrared signal to obtain a time-frequency spectrum diagram;

[0051] The time-frequency spectrum analysis model expression is:

[0052]

[0053] Among them, C s (t,f) is the time-frequency spectrum, f is frequency, t is time, f(ξ,τ) represents the kernel function; s(t) is the input near-infrared signal, τ is the time shift, ξ is the frequency shift, u is the integral variable, exp is the exponential function, σ is a constant, and * indicates conjugation;

[0054] S522, selecting 8 observation planes on the time-frequency spectrum diagram;

[0055] S523, obtaining a point set of contour lines on the eight observation planes;

[0056] S524, dividing the effective frequency range of the time-frequency spectrum into M segments, counting the number of intersections between each segment and the point set of the contour line according to frequency, and obtaining feature vector information, where M is an integer;

[0057] S525, processing the characteristic vector information to obtain a peak frequency, a rising rate, and a decay rate;

[0058] S526, the peak frequency, rising rate and decay rate constitute near infrared signal characteristic information.

[0059] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0060] (1) The present invention can simultaneously collect EEG signals and near-infrared signals, realizing the fusion of multimodal information, which helps to more comprehensively understand the brain's perception and cognitive process of olfactory stimulation.

[0061] (2) The present invention adopts high-precision synchronization control technology to ensure the temporal consistency of EEG signals and near-infrared signals, thereby improving the reliability and accuracy of the data.

[0062] (3) The system designed by the present invention takes into account the user's comfort and portability, facilitates experiments in different scenarios, and expands the scope of application. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0064] Figure 1It is a structural schematic diagram of an olfactory electroencephalogram signal recognition system disclosed in an embodiment of the present invention;

[0065] Figure 2 is a schematic diagram of an EEG signal acquisition module disclosed in an embodiment of the present invention;

[0066] Figure 3 is a schematic diagram of a near infrared signal acquisition module disclosed in an embodiment of the present invention;

[0067] Figure 4 is a schematic diagram of a synchronous control module disclosed in an embodiment of the present invention;

[0068] Figure 5 It is a flowchart of a method for identifying olfactory electroencephalogram signals disclosed in an embodiment of the present invention;

[0069] Figure 6 It is a structural diagram of the olfactory EEG signal recognition model disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0070] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or equipment.

[0072] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0073] The present invention discloses an olfactory electroencephalogram signal recognition system and method, the system includes an olfactory stimulation generation module, an electroencephalogram signal acquisition module, a near-infrared signal acquisition module, a synchronization control module and a data analysis and processing module; in response to a user's input instruction, under the control of the synchronization control module, the olfactory stimulation generation module is used to generate a smell; the electroencephalogram signal acquisition module is used to acquire the user's electroencephalogram signal; the near-infrared signal acquisition module is used to acquire the user's near-infrared signal; the electroencephalogram signal and the near-infrared signal are integrated to obtain an olfactory electroencephalogram signal data set; the olfactory electroencephalogram signal data set is processed by the data analysis and processing module to obtain an olfactory electroencephalogram signal recognition result. The present invention can simultaneously acquire electroencephalogram signals and near-infrared signals, realizes the fusion of multimodal information, and helps to more comprehensively understand the brain's perception and cognitive process of olfactory stimulation. The following are detailed descriptions.

[0074] Embodiment 1

[0075] See also Figure 1 , Figure 1 Schematic diagram of the structure of an olfactory EEG signal recognition system disclosed in an embodiment of the present invention. Figure 1 The described olfactory EEG signal recognition system is applied in the field of neuroscience and biomedical engineering technology, and can simultaneously and accurately collect EEG signals and near-infrared signals, providing more comprehensive and reliable data for olfactory neuroscience research, which is not limited in the embodiments of the present invention. Figure 1 As shown, the olfactory EEG signal recognition system includes an olfactory stimulation generation module, an EEG signal acquisition module, a near-infrared signal acquisition module, a synchronization control module and a data analysis and processing module;

[0076] The olfactory stimulation generation module is data-connected to the synchronization control module and is used to generate odor;

[0077] The olfactory stimulation generation module can precisely control the release time, concentration and duration of the odor.

[0078] The olfactory stimulation generation module stores a variety of different odor substances and selects and allocates them through the synchronous control module.

[0079] The EEG signal acquisition module is data-connected with the synchronization control module and the data analysis and processing module, and is used to acquire the user's EEG signal and send the EEG signal to the data analysis and processing module;

[0080] The EEG signal acquisition module is an EEG head-mounted device composed of two electrodes. The electrodes are connected to the preamplifier to amplify the weak EEG signals and convert them into digital signals through an analog-to-digital converter.

[0081] The near infrared signal acquisition module is data-connected with the synchronization control module and the data analysis and processing module, and is used to acquire the user's near infrared signal and send the near infrared signal to the data analysis and processing module;

[0082] The near-infrared signal acquisition module consists of two light sources and a detector. The light source emits near-infrared light, and the detector receives the light signal after being scattered and absorbed by brain tissue. Continuous wave near-infrared spectroscopy technology is used to measure the concentration changes of oxygenated hemoglobin and deoxygenated hemoglobin to obtain near-infrared signals.

[0083] The synchronization control module is data-connected with the olfactory stimulation generation module, the EEG signal acquisition module and the near-infrared signal acquisition module for data synchronization;

[0084] The synchronization control module is responsible for coordinating the work of the olfactory stimulation generation module, the EEG signal acquisition module, and the near-infrared signal acquisition module to ensure the precise synchronization of the three in time. The synchronization control module realizes synchronization control through the hardware synchronization interface and the software control program.

[0085] The data analysis and processing module is data-connected to the EEG signal acquisition module and the near-infrared signal acquisition module, and is used to process the EEG signal and the near-infrared signal to obtain an olfactory EEG signal recognition result.

[0086] Optionally, the olfactory stimulation generating module includes an odor generator and an odor delivery pipeline;

[0087] The odor generator is data-connected to the odor delivery pipeline and the synchronous control module, and is used to generate different odors under the control of the synchronous control module, and release the odors through the odor delivery pipeline;

[0088] The odor delivery pipeline is data-connected with the odor generator for outputting the odor.

[0089] Optionally, the synchronization control module includes an input unit, an odor control unit and a synchronization control unit;

[0090] The input unit is data-connected to the odor control unit and is used to input user instructions;

[0091] The odor control unit is data-connected to the odor generator, and is used to control the odor generator to produce different odors according to the user instructions;

[0092] The synchronization control unit is data-connected with the odor generator, the EEG signal acquisition module and the near-infrared signal acquisition module for data synchronization.

[0093] Figure 2is a schematic diagram of an EEG signal acquisition module disclosed in an embodiment of the present invention; Figure 3 is a schematic diagram of a near infrared signal acquisition module disclosed in an embodiment of the present invention; Figure 4 It is a schematic diagram of a synchronous control module disclosed in an embodiment of the present invention.

[0094] It can be seen that the present invention can collect EEG signals and near-infrared signals at the same time, realize the fusion of multimodal information, and help to more comprehensively understand the brain's perception and cognitive process of olfactory stimulation. The present invention adopts high-precision synchronization control technology to ensure the temporal consistency of EEG signals and near-infrared signals, thereby improving the reliability and accuracy of the data. The system designed by the present invention takes into account the comfort and portability of the user, facilitates experiments in different scenarios, and expands the scope of application.

[0095] Embodiment 2

[0096] See also Figure 5 , Figure 5 : is a flow chart of a method for identifying olfactory EEG signals disclosed in an embodiment of the present invention. Figure 5 The described olfactory EEG signal recognition method is applied to the fields of neuroscience and biomedical engineering technology, and can simultaneously and accurately collect EEG signals and near-infrared signals, providing more comprehensive and reliable data for olfactory neuroscience research, which is not limited in the embodiments of the present invention. Figure 5 As shown, the olfactory EEG signal recognition method includes:

[0097] S1, in response to a user's input instruction, under the control of a synchronous control module, using an olfactory stimulation generating module to generate an odor;

[0098] Specifically, according to the user's input instructions, the olfactory stimulation module generates an odor, and after the user smells the odor, his or her EEG signal and near-infrared signal will change.

[0099] S2, collecting the user's EEG signal using an EEG signal collection module;

[0100] S3, using a near infrared signal acquisition module to collect a near infrared signal of the user;

[0101] S4, integrating the EEG signal and the near infrared signal to obtain an olfactory EEG signal data set;

[0102] The integration is to construct an EEG signal subset and a near-infrared signal subset, and the EEG signal subset and the near-infrared signal subset constitute an olfactory EEG signal data set;

[0103] S5, using a data analysis and processing module to process the olfactory EEG signal data set to obtain an olfactory EEG signal recognition result.

[0104] Optionally, the data analysis and processing module is used to process the olfactory EEG signal data set to obtain an olfactory EEG signal recognition result, including:

[0105] S51, extracting features from the EEG signal to obtain EEG signal feature information;

[0106] S52, extracting features from the near infrared signal to obtain feature information of the near infrared signal;

[0107] S53, performing feature fusion on the EEG signal feature information and the near infrared signal feature information to obtain fused feature information;

[0108] The method of feature fusion is:

[0109] Embed the feature information of the EEG signal into Euclidean space to obtain Euclidean feature information;

[0110] Perform hyperbolic spatial embedding representation on the feature information of the near infrared signal to obtain the hyperbolic feature information;

[0111] Fusing the Euclidean feature information and the hyperbolic feature information to obtain fused feature information;

[0112] The fusion method is:

[0113] Y=αY1+βY2

[0114] Y is the fused feature information, Y1 is the Euclidean feature information, Y2 is the hyperbolic feature information, and α+β=1.

[0115] Euclidean space network embedding: Each network can be represented by two matrices: the adjacency matrix A and the node feature matrix X (if the node has no features, it can be the identity matrix), where each row x in X i Represents the characteristics of the node. First, normalize the adjacency matrix A:

[0116]

[0117] Where I is the identity matrix, is a diagonal matrix, D ii =∑ j A ij , A ij is the data of the i-th row and j-th column of the adjacency matrix A, D ii is the i-th element of matrix D.

[0118] In order to obtain the network structure representation in Euclidean space, it is necessary to perform graph convolution on the input feature matrix, which follows the following feedforward pass:

[0119]

[0120] in is the Euclidean space parameter matrix of the specified layer; σ(·) is a nonlinear function, such as ReLU(·)=max(0,·); is the node embedding matrix of the lth layer, the input layer Z 0 =X; d is the dimension of each node embedding. This feed-forward transfer enables each node to obtain the information of its neighbor nodes through the normalized adjacency matrix, thereby learning the network structure.

[0121] Hyperbolic space network embedding: The initial node features belong to the Euclidean space, which is first mapped to the hyperbolic space. First, use express The origin in <o,(0,x i )>=0,(0,x i ) can be considered as A point in is mapped to the hyperbolic space as follows :

[0122] H 0 =exp o (X)

[0123] exp o () represents the mapping operation. For a given network, a (l+1)-layer graph convolution feed-forward pass to generate a node embedding matrix in hyperbolic space is as follows:

[0124]

[0125] in is the hyperbolic space parameter matrix of the lth layer, is a hyperbolic transformation, AGG(·) is a neighborhood aggregation operation in a hyperbolic space, is a hyperbolic nonlinear activation function, H l is the output result of the lth layer.

[0126] Optionally, the method of performing feature fusion on the EEG signal feature information and the near infrared signal feature information to obtain fused feature information is:

[0127] Integrating the EEG signal feature information and the near infrared signal feature information to obtain fusion parameter information;

[0128] The fusion parameter information expression is:

[0129]

[0130] Among them, Z is the fusion parameter information, Z1 is the EEG signal feature information, Z2 is the near-infrared signal feature information, α1 and a2 are weight coefficients, α1+a2=1;

[0131] Centralizing the fusion parameter information to obtain centralized parameter information;

[0132] The centralization parameter information expression is:

[0133] Z1=Zu

[0134] Among them, Z1 is the centralized parameter information, z i is the i-th element in Z, with a total of n elements;

[0135] Performing covariance calculation on the centralized parameter information to obtain a covariance matrix;

[0136] Performing eigenvalue decomposition on the covariance matrix to obtain m eigenvalues;

[0137] Sort the m eigenvalues ​​and extract the eigenvectors corresponding to the first k eigenvalues;

[0138] Using the eigenvector reconstruction model, the eigenvectors corresponding to the first k eigenvalues ​​are processed to obtain fusion parameter information;

[0139] The feature vector reconstruction model expression is:

[0140] U=[p1Z1,p2Z1,…,p k Z1]

[0141] Among them, U is the fusion parameter information, [p1,p2,…,p k ] is the eigenvector corresponding to the first k eigenvalues.

[0142] S54, using the fusion feature information, training the olfactory EEG signal recognition model to obtain an optimized olfactory EEG signal recognition model;

[0143] The olfactory EEG signal recognition model is a model that combines Radon transform with an improved convolutional neural network algorithm. First, the fused feature information is preprocessed to obtain the Radon feature map, and then input into the improved convolutional neural network for training. The improved convolutional neural network model consists of 4 convolution modules 1, 2 convolution modules 2, 2 residual modules, a global average pooling layer, a Dropout layer, a fully connected layer, a Softmax layer, and a classification output layer. Figure 6As shown. Convolution module 1 is composed of convolution layer, batch normalization layer, ReLU layer, and maximum pooling layer connected in sequence. The convolution layer settings in the four convolution modules 1 are: 32 and 64 pixels of 5×5, the convolution kernel of the moving step is 2, and 128 and 256 pixels of 3×3, the convolution kernel of the moving step is 1. The purpose of using the ReLU function as the activation function in convolution module 1 is that it can not only prevent the network gradient from disappearing, but also has a fast calculation speed, which is conducive to network convergence.

[0144] In the deep network training, convolution module 2 and residual module are used for feature extraction. Convolution module 2 consists of convolution layer, batch normalization layer and Swish layer; residual module is composed of 2 convolution layers, batch normalization layer, Swish layer and jump layer connection combination to achieve feature fusion. The convolution layers in convolution module 2 and residual module are composed of 512 pixels of 3×3 and a moving step of 1. Finally, the purpose of combining the global average pooling (GAP) layer and the Dropout layer before the fully connected layer is to further improve the recognition and classification ability of the model.

[0145] S55, using the optimized olfactory electroencephalogram signal recognition model, processing the electroencephalogram signal and the near-infrared signal to be processed to obtain an olfactory electroencephalogram signal recognition result.

[0146] Optionally, the extracting features of the EEG signal to obtain EEG signal feature information includes:

[0147] S511, preprocessing the EEG signal to obtain a preprocessed EEG signal;

[0148] Preprocessing includes removing noise and artifacts.

[0149] The removal of noise and artifacts is achieved by a low-pass filtering method, which is not limited in this embodiment.

[0150] S512, performing first feature extraction on the preprocessed EEG signal to obtain first feature information;

[0151] S513, performing second feature extraction on the preprocessed EEG signal to obtain second feature information;

[0152] S514, extracting a third feature from the preprocessed EEG signal to obtain third feature information;

[0153] S515: Fusing the first feature information, the second feature information and the third feature information to obtain EEG signal feature information.

[0154] Optionally, the calculation formula of the first characteristic information is:

[0155]

[0156] Among them, H1 is the first feature information, The preprocessed EEG signal is X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

[0157] Optionally, the calculation formula of the second characteristic information is:

[0158]

[0159] Among them, H2 is the second characteristic information, L is the preset length, j = 1, 2, ..., L, δ j is the singular value of matrix D, matrix The preprocessed EEG signal is X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

[0160] Optionally, the calculation formula of the third characteristic information is:

[0161]

[0162] Among them, H3 is the third characteristic information, S i For x i The Fourier transform of i For S i The center frequency, X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

[0163] The method of fusing the first feature information, the second feature information and the third feature information to obtain the electroencephalogram signal feature information is the same as the method of fusing the electroencephalogram signal feature information and the near-infrared signal feature information to obtain the fused feature information.

[0164] The first feature information and the second feature information are integrated to obtain fourth feature information;

[0165] The third characteristic information and the fourth characteristic information are fused to obtain the EEG signal characteristic information.

[0166] Optionally, extracting features from the near infrared signal to obtain feature information of the near infrared signal includes:

[0167] S521, using a time-frequency spectrum analysis model, performing time-frequency processing on the near-infrared signal to obtain a time-frequency spectrum diagram;

[0168] The time-frequency spectrum analysis model expression is:

[0169]

[0170] Among them, C s (t,f) is the time-frequency spectrum, f is frequency, t is time, f(ξ,τ) represents the kernel function; s(t) is the input near-infrared signal, τ is the time shift, ξ is the frequency shift, u is the integral variable, exp is the exponential function, σ is a constant, and * indicates conjugation;

[0171] Optionally, after obtaining the time-spectrogram, the time-spectrogram may be denoised to obtain a denoised time-spectrogram, and the denoised time-spectrogram is input into S522 to replace the time-spectrogram therein for subsequent processing, wherein the step of denoising the time-spectrogram to obtain the denoised time-spectrogram includes:

[0172] The denoising neural network DnCNN is constructed, which consists of 17 modules including 3×3 convolutional layers, batch normalization layers, and activation layers connected in series; the denoising autoencoder DAE is constructed, and the forward propagation process of DAE includes two stages: encoding and decoding. During the encoding process, the size of the feature map gradually decreases and the number of channels gradually increases. During the decoding process, the decoder reconstructs the input image based on the feature map encoding. The size of the feature map gradually increases to the size of the input time-frequency image. Depending on the number of channels in the input image, the number of channels is reduced to 1 or 3 layer by layer.

[0173] The time-frequency spectrum is input into the first DnCNN, the second DnCNN and the DAE. The three branches extract the noise information of the input image at different scales respectively, and the three feature maps obtained are spliced ​​together in the channel dimension at the end of the branch. Finally, a convolutional layer is used to perform feature fusion to obtain the noise component in the input image. The denoised image (i.e., denoised time-frequency spectrum) can be obtained by subtracting the noise component from the input image.

[0174] S522, selecting 8 observation planes on the time-frequency spectrum diagram;

[0175] S523, obtaining a point set of contour lines on the eight observation planes;

[0176] S524, dividing the effective frequency range of the time-frequency spectrum into M segments, counting the number of intersections between each segment and the point set of the contour line according to frequency, and obtaining feature vector information, where M is an integer;

[0177] S525, processing the characteristic vector information to obtain a peak frequency, a rising rate, and a decay rate;

[0178] S526, the peak frequency, rising rate and decay rate constitute near infrared signal characteristic information.

[0179] It can be seen that the present invention can collect EEG signals and near-infrared signals at the same time, realize the fusion of multimodal information, and help to more comprehensively understand the brain's perception and cognitive process of olfactory stimulation. The present invention adopts high-precision synchronization control technology to ensure the temporal consistency of EEG signals and near-infrared signals, thereby improving the reliability and accuracy of the data. The system designed by the present invention takes into account the comfort and portability of the user, facilitates experiments in different scenarios, and expands the scope of application.

[0180] Embodiment 3

[0181] In this embodiment, the user wears an EEG cap (EEG signal acquisition module) and a near-infrared probe (near-infrared signal acquisition module) and sits in a comfortable seat. The synchronous control module controls the olfactory stimulation generation module to release specific odor stimulation, and at the same time, the synchronous control module triggers the EEG signal acquisition module and the near-infrared signal acquisition module to start collecting signals. The collected data is transmitted to the data analysis and processing module for processing and analysis, and finally the EEG event-related potential and near-infrared hemodynamic response related to the olfactory stimulation are obtained.

[0182] Embodiment 4

[0183] Healthy volunteers were selected as subjects for the olfactory stimulation experiment. Before the experiment, the acquisition device was calibrated and debugged to ensure its stable performance. After the subjects put on the EEG cap and near-infrared probe, they adapted to the experimental environment for 10 minutes. During the experiment, three different odor stimuli were randomly released, and each odor stimulus was repeated 10 times, with an interval of 30 seconds between each stimulation. After processing and analysis of the collected data, the EEG event-related potentials and near-infrared hemodynamic responses related to each odor stimulus were obtained. The results showed that there were significant differences in brain responses caused by different odor stimuli.

[0184] Embodiment 5

[0185] Patients with olfactory disorders were tested and compared with healthy controls. The results showed that the EEG activity and hemodynamic changes in certain brain regions of the patient group were significantly different from those of the healthy control group, providing valuable reference for the diagnosis and treatment of the disease.

[0186] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0187] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0188] Finally, it should be noted that the olfactory EEG signal recognition system and method disclosed in the embodiments of the present invention disclose only the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An olfactory electroencephalogram signal recognition system, characterized in that: The system includes an olfactory stimulation generation module, an EEG signal acquisition module, a near-infrared signal acquisition module, a synchronization control module and a data analysis and processing module; The olfactory stimulation generation module is data-connected to the synchronization control module and is used to generate odor; The EEG signal acquisition module is data-connected with the synchronization control module and the data analysis and processing module, and is used to acquire the user's EEG signal and send the EEG signal to the data analysis and processing module; The near infrared signal acquisition module is data-connected with the synchronization control module and the data analysis and processing module, and is used to acquire the near infrared signal of the user and send the near infrared signal to the data analysis and processing module; The synchronization control module is data-connected with the olfactory stimulation generation module, the EEG signal acquisition module and the near-infrared signal acquisition module for data synchronization; The data analysis and processing module is data-connected to the EEG signal acquisition module and the near-infrared signal acquisition module, and is used to process the EEG signal and the near-infrared signal to obtain an olfactory EEG signal recognition result.

2. The olfactory EEG signal recognition system according to claim 1, characterized in that: The olfactory stimulation generation module includes an odor generator and an odor delivery pipeline; The odor generator is data-connected to the odor delivery pipeline and the synchronous control module, and is used to generate different odors under the control of the synchronous control module, and release the odors through the odor delivery pipeline; The odor delivery pipeline is data-connected with the odor generator for outputting the odor.

3. The olfactory EEG signal recognition system according to claim 1, characterized in that: The synchronization control module includes an input unit, an odor control unit and a synchronization control unit; The input unit is data-connected to the odor control unit and is used to input user instructions; The odor control unit is data-connected to the odor generator, and is used to control the odor generator to produce different odors according to the user instructions; The synchronization control unit is data-connected with the odor generator, the EEG signal acquisition module and the near-infrared signal acquisition module for data synchronization.

4. A method for identifying olfactory electroencephalogram signals, characterized in that: The method is applicable to the olfactory electroencephalogram signal recognition system according to any one of claims 1 to 3, and the method comprises: S1, in response to a user's input instruction, under the control of a synchronous control module, using an olfactory stimulation generating module to generate an odor; S2, collecting the user's EEG signal using an EEG signal collection module; S3, using a near infrared signal acquisition module to collect a near infrared signal of the user; S4, integrating the EEG signal and the near infrared signal to obtain an olfactory EEG signal data set; S5, using a data analysis and processing module to process the olfactory EEG signal data set to obtain an olfactory EEG signal recognition result.

5. The method for identifying olfactory electroencephalogram signals according to claim 4, characterized in that: The data analysis and processing module is used to process the olfactory EEG signal data set to obtain an olfactory EEG signal recognition result, including: S51, extracting features from the EEG signal to obtain EEG signal feature information; S52, extracting features from the near infrared signal to obtain feature information of the near infrared signal; S53, performing feature fusion on the EEG signal feature information and the near infrared signal feature information to obtain fused feature information; S54, using the fusion feature information, training the olfactory EEG signal recognition model to obtain an optimized olfactory EEG signal recognition model; S55, using the optimized olfactory electroencephalogram signal recognition model, processing the electroencephalogram signal and the near-infrared signal to be processed to obtain an olfactory electroencephalogram signal recognition result.

6. The method for identifying olfactory electroencephalogram signals according to claim 5, characterized in that: The step of extracting features from the EEG signal to obtain EEG signal feature information includes: S511, preprocessing the EEG signal to obtain a preprocessed EEG signal; S512, performing first feature extraction on the preprocessed EEG signal to obtain first feature information; S513, performing second feature extraction on the preprocessed EEG signal to obtain second feature information; S514, extracting a third feature from the preprocessed EEG signal to obtain third feature information; S515: Fusing the first feature information, the second feature information and the third feature information to obtain EEG signal feature information.

7. The method for identifying olfactory electroencephalogram signals according to claim 6, characterized in that: The calculation formula of the first characteristic information is: Among them, H1 is the first feature information, The preprocessed EEG signal is X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

8. The method for identifying olfactory electroencephalogram signals according to claim 6, characterized in that: The calculation formula of the second characteristic information is: Among them, H2 is the second characteristic information, L is the preset length, j = 1, 2, ..., L, δ j is the singular value of matrix D, matrix The preprocessed EEG signal is X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

9. The method for identifying olfactory electroencephalogram signals according to claim 6, characterized in that: The calculation formula of the third characteristic information is: Among them, H3 is the third characteristic information, S i For x i The Fourier transform of i For S i The center frequency, X = {x i ,i=1,2,…,N}, N is the signal length, x i is the i-th signal sample point.

10. The method for identifying olfactory electroencephalogram signals according to claim 5, characterized in that: The step of extracting features from the near infrared signal to obtain feature information of the near infrared signal includes: S521, using a time-frequency spectrum analysis model, performing time-frequency processing on the near-infrared signal to obtain a time-frequency spectrum diagram; The time-frequency spectrum analysis model expression is: Among them, C s (t,f) is the time-frequency spectrum, f is frequency, t is time, f(ξ,τ) represents the kernel function; s(t) is the input near-infrared signal, τ is the time shift, ξ is the frequency shift, u is the integral variable, exp is the exponential function, σ is a constant, and * indicates conjugation; S522, selecting 8 observation planes on the time-frequency spectrum diagram; S523, obtaining a point set of contour lines on the eight observation planes; S524, dividing the effective frequency range of the time-frequency spectrum into M segments, counting the number of intersections between each segment and the point set of the contour line according to frequency, and obtaining feature vector information, where M is an integer; S525, processing the characteristic vector information to obtain a peak frequency, a rising rate, and a decay rate; S526, the peak frequency, rising rate and decay rate constitute near infrared signal characteristic information.

Images